Merge tag 'nfs-for-4.18-2' of git://git.linux-nfs.org/projects/trondmy/linux-nfs

Pull NFS client bugfixes from Trond Myklebust:
 "Hightlights include:

   - fix an rcu deadlock in nfs_delegation_find_inode()

   - fix NFSv4 deadlocks due to not freeing the session slot in
     layoutget

   - don't send layoutreturn if the layout is already invalid

   - prevent duplicate XID allocation

   - flexfiles: Don't tie up all the rpciod threads in resends"

* tag 'nfs-for-4.18-2' of git://git.linux-nfs.org/projects/trondmy/linux-nfs:
  pNFS/flexfiles: Process writeback resends from nfsiod context as well
  pNFS/flexfiles: Don't tie up all the rpciod threads in resends
  sunrpc: Prevent duplicate XID allocation
  pNFS: Don't send layoutreturn if the layout is already invalid
  pNFS: Always free the session slot on error in nfs4_layoutget_handle_exception
  NFS: Fix an rcu deadlock in nfs_delegation_find_inode()

6 files changed, 59 insertions, 16 deletions

diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c
index bbd0465535eb..f033f3a69a3b 100644
--- a/fs/nfs/delegation.c
+++ b/fs/nfs/delegation.c
@@ -883,8 +883,10 @@ struct inode *nfs_delegation_find_inode(struct nfs_client *clp,
 	rcu_read_lock();
 	list_for_each_entry_rcu(server, &clp->cl_superblocks, client_link) {
 		res = nfs_delegation_find_inode_server(server, fhandle);
-		if (res != ERR_PTR(-ENOENT))
+		if (res != ERR_PTR(-ENOENT)) {
+			rcu_read_unlock();
 			return res;
+		}
 	}
 	rcu_read_unlock();
 	return ERR_PTR(-ENOENT);
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index d4a07acad598..8f003792ccde 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -1243,17 +1243,18 @@ static int ff_layout_read_done_cb(struct rpc_task *task,
 					   hdr->ds_clp, hdr->lseg,
 					   hdr->pgio_mirror_idx);
 
+	clear_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags);
+	clear_bit(NFS_IOHDR_RESEND_MDS, &hdr->flags);
 	switch (err) {
 	case -NFS4ERR_RESET_TO_PNFS:
 		if (ff_layout_choose_best_ds_for_read(hdr->lseg,
 					hdr->pgio_mirror_idx + 1,
 					&hdr->pgio_mirror_idx))
 			goto out_eagain;
-		ff_layout_read_record_layoutstats_done(task, hdr);
-		pnfs_read_resend_pnfs(hdr);
+		set_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags);
 		return task->tk_status;
 	case -NFS4ERR_RESET_TO_MDS:
-		ff_layout_reset_read(hdr);
+		set_bit(NFS_IOHDR_RESEND_MDS, &hdr->flags);
 		return task->tk_status;
 	case -EAGAIN:
 		goto out_eagain;
@@ -1403,6 +1404,10 @@ static void ff_layout_read_release(void *data)
 	struct nfs_pgio_header *hdr = data;
 
 	ff_layout_read_record_layoutstats_done(&hdr->task, hdr);
+	if (test_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags))
+		pnfs_read_resend_pnfs(hdr);
+	else if (test_bit(NFS_IOHDR_RESEND_MDS, &hdr->flags))
+		ff_layout_reset_read(hdr);
 	pnfs_generic_rw_release(data);
 }
 
@@ -1423,12 +1428,14 @@ static int ff_layout_write_done_cb(struct rpc_task *task,
 					   hdr->ds_clp, hdr->lseg,
 					   hdr->pgio_mirror_idx);
 
+	clear_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags);
+	clear_bit(NFS_IOHDR_RESEND_MDS, &hdr->flags);
 	switch (err) {
 	case -NFS4ERR_RESET_TO_PNFS:
-		ff_layout_reset_write(hdr, true);
+		set_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags);
 		return task->tk_status;
 	case -NFS4ERR_RESET_TO_MDS:
-		ff_layout_reset_write(hdr, false);
+		set_bit(NFS_IOHDR_RESEND_MDS, &hdr->flags);
 		return task->tk_status;
 	case -EAGAIN:
 		return -EAGAIN;
@@ -1575,6 +1582,10 @@ static void ff_layout_write_release(void *data)
 	struct nfs_pgio_header *hdr = data;
 
 	ff_layout_write_record_layoutstats_done(&hdr->task, hdr);
+	if (test_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags))
+		ff_layout_reset_write(hdr, true);
+	else if (test_bit(NFS_IOHDR_RESEND_MDS, &hdr->flags))
+		ff_layout_reset_write(hdr, false);
 	pnfs_generic_rw_release(data);
 }
 
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index ed45090e4df6..6dd146885da9 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -3294,6 +3294,7 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 	struct nfs4_closedata *calldata = data;
 	struct nfs4_state *state = calldata->state;
 	struct inode *inode = calldata->inode;
+	struct pnfs_layout_hdr *lo;
 	bool is_rdonly, is_wronly, is_rdwr;
 	int call_close = 0;
 
@@ -3337,6 +3338,12 @@ static void nfs4_close_prepare(struct rpc_task *task, void *data)
 		goto out_wait;
 	}
 
+	lo = calldata->arg.lr_args ? calldata->arg.lr_args->layout : NULL;
+	if (lo && !pnfs_layout_is_valid(lo)) {
+		calldata->arg.lr_args = NULL;
+		calldata->res.lr_res = NULL;
+	}
+
 	if (calldata->arg.fmode == 0)
 		task->tk_msg.rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_CLOSE];
 
@@ -5972,12 +5979,19 @@ static void nfs4_delegreturn_release(void *calldata)
 static void nfs4_delegreturn_prepare(struct rpc_task *task, void *data)
 {
 	struct nfs4_delegreturndata *d_data;
+	struct pnfs_layout_hdr *lo;
 
 	d_data = (struct nfs4_delegreturndata *)data;
 
 	if (!d_data->lr.roc && nfs4_wait_on_layoutreturn(d_data->inode, task))
 		return;
 
+	lo = d_data->args.lr_args ? d_data->args.lr_args->layout : NULL;
+	if (lo && !pnfs_layout_is_valid(lo)) {
+		d_data->args.lr_args = NULL;
+		d_data->res.lr_res = NULL;
+	}
+
 	nfs4_setup_sequence(d_data->res.server->nfs_client,
 			&d_data->args.seq_args,
 			&d_data->res.seq_res,
@@ -8650,6 +8664,8 @@ nfs4_layoutget_handle_exception(struct rpc_task *task,
 
 	dprintk("--> %s tk_status => %d\n", __func__, -task->tk_status);
 
+	nfs4_sequence_free_slot(&lgp->res.seq_res);
+
 	switch (nfs4err) {
 	case 0:
 		goto out;
@@ -8714,7 +8730,6 @@ nfs4_layoutget_handle_exception(struct rpc_task *task,
 		goto out;
 	}
 
-	nfs4_sequence_free_slot(&lgp->res.seq_res);
 	err = nfs4_handle_exception(server, nfs4err, exception);
 	if (!status) {
 		if (exception->retry)
@@ -8786,20 +8801,22 @@ nfs4_proc_layoutget(struct nfs4_layoutget *lgp, long *timeout)
 	if (IS_ERR(task))
 		return ERR_CAST(task);
 	status = rpc_wait_for_completion_task(task);
-	if (status == 0) {
+	if (status != 0)
+		goto out;
+
+	/* if layoutp->len is 0, nfs4_layoutget_prepare called rpc_exit */
+	if (task->tk_status < 0 || lgp->res.layoutp->len == 0) {
 		status = nfs4_layoutget_handle_exception(task, lgp, &exception);
 		*timeout = exception.timeout;
-	}
-
+	} else
+		lseg = pnfs_layout_process(lgp);
+out:
 	trace_nfs4_layoutget(lgp->args.ctx,
 			&lgp->args.range,
 			&lgp->res.range,
 			&lgp->res.stateid,
 			status);
 
-	/* if layoutp->len is 0, nfs4_layoutget_prepare called rpc_exit */
-	if (status == 0 && lgp->res.layoutp->len)
-		lseg = pnfs_layout_process(lgp);
 	rpc_put_task(task);
 	dprintk("<-- %s status=%d\n", __func__, status);
 	if (status)
@@ -8817,6 +8834,8 @@ nfs4_layoutreturn_prepare(struct rpc_task *task, void *calldata)
 			&lrp->args.seq_args,
 			&lrp->res.seq_res,
 			task);
+	if (!pnfs_layout_is_valid(lrp->args.layout))
+		rpc_exit(task, 0);
 }
 
 static void nfs4_layoutreturn_done(struct rpc_task *task, void *calldata)
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index a8f5e6b16749..3fe81424337d 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -801,6 +801,11 @@ static inline void nfs4_lgopen_release(struct nfs4_layoutget *lgp)
 {
 }
 
+static inline bool pnfs_layout_is_valid(const struct pnfs_layout_hdr *lo)
+{
+	return false;
+}
+
 #endif /* CONFIG_NFS_V4_1 */
 
 #if IS_ENABLED(CONFIG_NFS_V4_2)
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 9dee3c23895d..712eed156d09 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -1438,6 +1438,8 @@ enum {
 	NFS_IOHDR_EOF,
 	NFS_IOHDR_REDO,
 	NFS_IOHDR_STAT,
+	NFS_IOHDR_RESEND_PNFS,
+	NFS_IOHDR_RESEND_MDS,
 };
 
 struct nfs_io_completion;
diff --git a/net/sunrpc/xprt.c b/net/sunrpc/xprt.c
index 3c85af058227..3fabf9f6a0f9 100644
--- a/net/sunrpc/xprt.c
+++ b/net/sunrpc/xprt.c
@@ -987,8 +987,6 @@ bool xprt_prepare_transmit(struct rpc_task *task)
 		task->tk_status = -EAGAIN;
 		goto out_unlock;
 	}
-	if (!bc_prealloc(req) && !req->rq_xmit_bytes_sent)
-		req->rq_xid = xprt_alloc_xid(xprt);
 	ret = true;
 out_unlock:
 	spin_unlock_bh(&xprt->transport_lock);
@@ -1298,7 +1296,12 @@ void xprt_retry_reserve(struct rpc_task *task)
 
 static inline __be32 xprt_alloc_xid(struct rpc_xprt *xprt)
 {
-	return (__force __be32)xprt->xid++;
+	__be32 xid;
+
+	spin_lock(&xprt->reserve_lock);
+	xid = (__force __be32)xprt->xid++;
+	spin_unlock(&xprt->reserve_lock);
+	return xid;
 }
 
 static inline void xprt_init_xid(struct rpc_xprt *xprt)
@@ -1316,6 +1319,7 @@ void xprt_request_init(struct rpc_task *task)
 	req->rq_task	= task;
 	req->rq_xprt    = xprt;
 	req->rq_buffer  = NULL;
+	req->rq_xid	= xprt_alloc_xid(xprt);
 	req->rq_connect_cookie = xprt->connect_cookie - 1;
 	req->rq_bytes_sent = 0;
 	req->rq_snd_buf.len = 0;